Yes, I can: Audio file loading slowly since Devialet RAAT
Yes, I did watch the video, painful as it was. He talked about jitter and he talked about network timing affecting the D/A converter. It doesnât - the TCP/IP data is buffered and checksummed at the receiving ethernet port on the endpoint. That data is then read from the buffer (think of it at this point as if you were reading the data from a drive) and sent out over USB/SPDIF/TOSLINK/AES/I2S to the DAC.
Jitter can occur from reading from the endpoint data buffer onwards, but none of that is caused by the network. Roon uses TCP/IP, not UDP, so the network timing has no effect on the endpoint/DAC.
This stuff is really very simple - I have no clue as to why people make it seem complicated.
2 Likes
Itâs funny to me that such a fuzz is made about (ethernet) network quality and its effect on audio. Apart from the fact that with TCP/IP there isnât any such effect, I really wonder why the people that THINK there is donât worry about jitter and other effects on their bank account or transactions.
After all, banks (and everything else) work with TCP/IP. And at a much, very much, higher frequency then our ears can master.
Iâm very sure the datacenters of banks donât use specialized cables.
2 Likes
Perhaps they have larger transactions to worry about 
1 Like
If something looks too simple, there must be a much deeper conspiracy behind it.
4 Likes
Because subjectivist audiophiles seemingly enjoy the proverbial equivalents of wearing hairshirts and flagellating themselves. Extreme faith. For high end audio to be as good as it gets requires suffering through complication and great expense. Anything that is easy to use/implement or is inexpensive must be compromised in sound quality.
AJ
8 Likes
âAny missing / corrupted packets are resentâ.
I suspect the timing of the arrivals and the inconsistency of the time windows between these packets may cause some impacts to the SQ.
Reading todays post seems like dejavu of my own thinking (TCP, buffers all around, 0.0 errors on all network interfaces etc - simply all is 100% ok from technical perspective, even on moderately higher level than average situations).
So there is simply no room for any problem for streamed audio until very significant issues occur (be it from NAS or internet).
I have probably similar IT background as many other people in this topic - like I wrote x86 assembly code back around '97 utilising half-duplex coaxial ethernet connection to establish semi-duplex audio communication between two nodes on the same ethernet segment (it was feature of broader-functionality communication system). So without a shame I would say I know something about streaming audio, but I do not know all
It took me years to realise things in audio are not always by the books. Or better said - they are, but we need to look into further books. Actually come down to lower levels and taking into account imperfections which will always be there, just masked to different level (like RFI polluted environment, or power line variations in populated areas, usage of nasty SMPS near audio etc). When all this imperfections add to each other, they simply cause distortion in reproduced audio signal which usually on low levels causes slight lose of resolution, or even making sound rough/glassy .
One of those imperfections is caused by network traffic - which in low level words means MHz/GHz signal transmission via ethernet cables (think of those cables radiating GHz spikes of some energy affecting mainly clocks in DAC e.g. jitter), then you have also streamerâs network interface&CPU power consumption and EMI radiation increased. Iâm sure there could more including noise propagation via signal conductors and grounds, but letâs now see actually what happens on CPU when there is a network traffic processed - there is so called hardware interrupt - which will stop CPU core from what it is doing (which will be during particular moments actually communicating with USB interface as part of streaming process) - this is generally fine as there are buffers for USB (and other) Audio, however this will affect energy/EMI variation profile of USB interface in DAC itself.
This is something i totally missed for quite a long time. But iâm not much sorry about that, because I had to solve few other issues to actually get system up to level of required transparency to be able to play with my digital streamer sw tuning experiments which is my hobby last few years (and yes, my doctor is fine with that 
).
The biggest variation is our systems and that is great, as we all have our own way we listen to them, or in some rare cases actually listen to music thru them 
4 Likes
It doesnât. The packets are reordered at the receiving end as they were transmitted. If the system canât keep up, you get dropped/lost packets and interruption to the sound. TCP/IP is all or nothing.
5 Likes
Cannot be repeated too often.
If it works (music is audible) no cable, extra re-gen can improve nothing at all. The capacity (transport of data packets per time) is so much overkill re audio purposes that the receiving end (streamer) has all time and more to resort or request re-transmission if necessary.
Plus, as a wise man here said "Donât use shielded ethernet cablesâ
And you are as good as it gets. Literally.
4 Likes
Shielded cables use is primary in industrial applications due to big EMI etc. while in SOHO environment not needed / recommended. But shielded cables can be used in SOHO when solving particular problem and when few basic conditions are met:
- cables are properly terminated
- networking HW is grounded (but this is usually not consumer grade because SOHO networking devices use just 2 power plug pins not 3 pins - 3rd for grounding)
- all networking devices (including audio HW) must use the same grounding point / earth potential
Anyway I definitely agree that in home conditions cannot be such big EMI to need use shielded cables. The only exception is roof ISP antenna which is PoE powered through shielded cable and grounded properly (PoE has always 3 pins power plug)
The guys at Wigwam just did a listening test with a few switches/boxes. Their conclusion, like many, was that it made a noteworthy difference. Rather than theorising/measuring, they listened. Old school I know butâŚ
Blind test?
1 Like
Do you have a link?
No, a listening test. You know, when you change things around and see if it make an audible difference?
Why not have a go, go wild, put a blindfold on, everything.
Because - mysteriously - if audiophiles are suddenly urged to judge by their ears only, these test results like you are citing collapse like a house of cards. Canât have it both ways. Audiophile and videophile âŚ
Ever wondered why none of these gurus ever accept a proper blind test to verify their skills? Wanna guess?
3 Likes
The test result donât collapse because you want them to?
When you start listening to audio with something other than your ears, letâs start using that to test?
Both ways? Not sure what youâre on about? Gurus? People who listen to hifi more like.
âproper blind testâ- Iâm not sure they are obsessed enough and probably donât find it necessary, why not do your own and report back? Blind/AB/BA/Naked/Clothed, go wild make up as many as you want, just have fun doing it.
The argument that has been repeated ad nauseum in this thread, to counter the claim that ethernet cables, routers and switches can affect sound quality, is that thereâs no objective explanation nor any acceptable evidence to suggest that this is this case. For you to point out that there are yet more people who have made the same erroneous assumption, i.e. that thereâs an objective causal relationship between changing a network component and actual rather than perceived SQ, doesnât move the discussion forward. It doesnât matter how many people believe something to be true if itâs false.
Yep, you can change things around, and Iâm absolutely 100% convinced that you do hear a difference, but this doesnât confirm any sort of objectively causal relationship. It just confirms that you perceive a difference. One that is most likely attributable to psychoacoustic effects or some form of expectation. Until you rule out that possibility âŚ
Because it would be a pointless waste of time. As @Bernd_Kurte said, itâs no surprise that not a single one of these gurus or manufactures of boutique network gear have conducted a proper blind test to confirm.
7 Likes
Well if you did, you could speak from authority and experience rather than just opinion. Whatâs to lose but your preconceptions?
The problem is, even if I heard a difference, I wouldnât believe my ears. So experience is, as usual, little use when put up against facts and physics. Authority? How does getting drunk (presumably I would have to) and indulging in nonsense confer authority?
1 Like